52results about How to "Realize automatic recommendation" patented technology

The invention discloses a data access method and device, computer equipment and a readable storage medium. At least one first field to be mapped corresponding to a first external dataset to be accessed is acquired; based on a term vector corresponding to each participle in the first fields to be mapped, a dataset vector to be matched corresponding to the first external dataset is obtained; based on the dataset vector to be matched and a dataset sorting model trained in advance, a first standard dataset matched with the first external dataset is determined; based on the similarity between standard fields in the first standard dataset and the first fields to be mapped, first field mapping relationships between the first fields to be mapped and the standard fields are built; based on the first field mapping relationships, the first external dataset is accessed into the first standard dataset. By means of the data access method and device, the computer equipment and the readable storage medium, automatic recommendation of data access and automatic mapping of field levels can be achieved.

The embodiment of the invention discloses an interior-design method and system based on user interest preference. The method includes: carrying out editing design of 3D design schemes by a designer for interior design, and uploading the 3D design schemes and interior-design 2D effect drawings, which are associated therewith, to a server; setting initial preference by a user, automatically recommending interior-design 2D effect drawings to a user mobile-end by the server according to the user preference; uploading an interest picture by the user to acquire 2D effect drawings associated with theinterest image and corresponding 3D design schemes; clicking an interior-design 2D effect drawing by the user to automatically acquire a presentation effect of a 3D design scheme; and communicating with the designer by the user to communicate the 3D design scheme online. According to the method, user attributes, browsed-content attributes and operation behaviors of the user in the system are simultaneously investigated, the user interest preference is effectively understood, automatic recommendation of the design schemes is realized, the user and the designer can simultaneously and jointly participate in scheme design online, communication costs are reduced, and efficiency of communication between the designer and the customer is improved.

The invention discloses a node cluster implementation method, device and system in a peer-to-peer network, which are used for realizing automatic promotion of a main node during peer-to-peer network cluster deployment, reducing the complexity of the cluster deployment and improving the efficiency of the cluster deployment. The method comprises the following steps: acquiring state information of each node with same cluster identifier by any node in a cluster, wherein the state information comprises the identifier and a performance parameter of the node; determining a main node according to the acquired performance parameter of each node and the performance parameter of the node.

The invention relates to an identification binding method and device, a terminal, a server and a storage medium, and belongs to the technical field of Internet. The method comprises the steps that based on a logged-in first user identification, an application interface of a game client side is displayed and comprises a face-to-face group building option; when a triggering operation on the face-to-face group building option is detected, an identification input field is displayed in the application interface, and a recognition identification input in the identification input field is acquired; group building requests are sent to the server and carry the recognition identification and the first user identification, the server is used for binding a game group identification generated accordingto the recognition identification and the first user identification carried in the group building requests when receiving the group building requests carrying the same recognition identification andthe corresponding first user identification. It is unnecessary to build an association relationship among user identifications, a restriction that the association relationship has to be built among the user identifications is got rid of, the flexibility is improved, and an application range is enlarged.

The invention relates to the field of voice recognition, and provides a product recommending method based on voice analysis. The method comprises the following steps of playing preset initial voice associated with a product inquiry request when receiving the product inquiry request; collecting first voice information replied by a customer; recognizing the first voice information to obtain the corresponding first tendency information; inquiring a preset verbal trick forest; obtaining a target node matched with the first tendency information in the preset verbal trick forest; generating and broadcasting first recommended verbal trick corresponding to the first tendency information according to the target node; receiving second voice information relied by the customer on the basis of the first recommended verbal trick; recognizing the second voice information to obtain the corresponding second tendency information; when the second tendency information is negative, performing backtrackingreaching on the target node according to the second tendency information; obtaining and broadcasting the second recommended verbal trick. The invention also discloses a product recommending device based on voice analysis, product recommending equipment based on voice analysis, and a storage medium. The product recommending accuracy is improved.

The invention relates to the technical field of intelligent air conditioners, and discloses a method, device and equipment for air conditioner working mode recommendation. The method comprises the steps that user use data information corresponding to each air conditioner of a set model within a set time period is obtained; index data information matched with the energy consumption of the air conditioner of the set model is extracted from the user use data information; an energy-saving comfortable working mode corresponding to the set model is determined through a clustering learning algorithmaccording to each piece of energy consumption information and the corresponding index data information; and the energy-saving comfortable working mode is recommended to the air conditioner of the setmodel. Thus, automatic recommendation of the energy-saving comfortable working mode is achieved, the dual requirements for energy saving and comfort of the user are met, and the intelligence of the air conditioner and the user experience are improved.

The invention discloses a cloud resource recommendation method and device, an electronic terminal and a storage medium. The cloud resource recommendation method comprises the steps of receiving a cloud resource recommendation instruction, analyzing a target demand parameter carried in the cloud resource recommendation instruction, determining the type of the target demand parameter, and obtainingan association relationship of demand-cloud resources corresponding to the type of the target demand parameter, wherein the association relationship of the demand-cloud resources is determined based on historical recommendation information of the cloud resources; and determining target cloud resource information matched with the target demand parameter according to the association relationship ofthe demand-cloud resources, and recommending the target cloud resource information. The target cloud resource information does not need to be determined manually, and automatic recommendation of the cloud resources is realized, so that low-resource-consumption and high-efficiency cloud resource recommendation is realized. Meanwhile, by analyzing the demand parameters of the user, the cloud demandof the user is fully understood, and high-accuracy cloud resource recommendation is realized.

The invention discloses an item handling intelligent recommendation method based on a data graph and a service graph. The method comprises the steps: S1, constructing a data graph body; S2, constructing a data graph; and S3, constructing a service map. After the S3, item recommendation can be further carried out based on the data graph and the service graph. After the S3, the method can further comprise: S4, updating the service graph data; and S5, recommending a potential item. By means of the item handling intelligent recommendation method based on the data graph and the service graph, iteminformation can be automatically recommended, the interaction process of a user and a system is reduced, and meanwhile the item recommendation accuracy is improved.

The invention discloses a program recommendation method, a program recommendation system and a computer readable storage medium, and the method comprises the following steps: when that a user watchesa program is sensed, acquiring a face image of a current user; performing feature extraction for the face image of the current user; matching the extracted face feature of the current user with the face feature prestored in a cloud database, wherein the face features of different users and corresponding program parameters are stored in the cloud database; if the corresponding face feature is matched in the cloud database, acquiring the program parameter corresponding to the face feature of the current user from the cloud database; and recommending a corresponding program to the current user according to the program parameter. The method and the system perform program recommendation for the user in combination with the face feature of the current user and the user face features and the program parameters stored in the cloud database, not only can effectively protect privacy information of the user in program recommendation, but also can efficiently and accurately realize automatic recommendation of programs, thereby improve user experience.

The invention relates to the field of artificial intelligence, and provides a course recommendation method and device based on unsupervised learning, equipment and a medium, which can adopt a cycle-GAN model to extend training data, effectively solve the problems of poor model training effect and poor precision caused by insufficient samples, and meanwhile, improve the accuracy of the training data due to the characteristics of the cycle-GAN model. In addition, the samples generated after expansion have high availability, and the recommended courses are more reasonable due to the fact that prediction results of different dimensions are integrated, so that the courses are automatically recommended. The invention also relates to a blockchain technology, and the involved model can be stored in the blockchain.

The invention relates to the technical field of intelligent equipment, and discloses a clothes fitting recommendation method, device and equipment. The method comprises the following steps: acquiringa to-be-dressed picture containing one or more to-be-dressed clothes; inputting the to-be-dressed picture into a configured multi-label single model to obtain a to-be-dressed clothes attribute featurevector; determining at least one to-be-recommended picture as a first to-be-recommended picture according to the acquired recommended clothing attribute feature vector of each to-be-recommended picture and the to-be-dressed clothing attribute feature vector, and recommending the first to-be-recommended picture. In this way, automatic clothes fitting recommendation is achieved, and the intelligence of equipment related to clothes is improved.

The invention provides a method and a device for recommending service. The method comprises the following steps: obtaining an identifier of a user; sending a first request to a second server, the first request being used for indicating the second server to obtain a service recommendation file stored in association with the identity file after determining that the identity file including the identifier is stored; wherein the service recommendation file is generated by the first server according to the service use data of the user and then sent to the second server, and the service use data is obtained by a deep packet inspection (DPI) server according to a service feature library and each first internet surfing data of the user within a first preset duration; and obtaining the service recommendation file from the second server, and sending service recommendation information to target equipment, the service recommendation information comprising a to-be-recommended service included in theservice recommendation file. According to the invention, the service can be accurately recommended to the user, and the service recommendation process can be automatically triggered when the user needs to do the service.

The invention belongs to the technical field of interior design. An interior design method and system based on user interest preference are disclosed. The system comprises a user mobile terminal, a server, a designer mobile terminal, a cloud database, a picture upload module, a personal center module, a 3D solution experience module, a management center module, an algorithm processor, a data processor, a designer personal center module, a 3D object editor, and a 3D solution generator. The mobile client is wirelessly connected to the server, the designer mobile end is wirelessly connected to the server, and the cloud database is wirelessly connected to the server. The invention inspects the user attribute, the attribute of the browsed content and the operation behavior of the user in the system, effectively understands the interest preference of the user, and realizes the automatic recommendation of the design scheme, and the user and the designer can participate in the scheme design simultaneously online, reduces the communication cost, and improves the communication efficiency between the designer and the customer.

Provided are an information recommending method, an electronic terminal, and a computer readable storage medium message. The information recommending method comprises: identifying a current motion mode of a user; based on feature points of reference information, searching target information matched with the feature points, wherein the reference information is arranged in advance and is associatedwith the motion mode which is identified currently; and outputting the target information, to recommend the target information to the user. The scheme provided by the application can realize information automatic recommendation based on the motion mode.

The invention discloses a recommendation method and device of ceramic design. The method includes the steps of obtaining a request for the ceramic design, wherein the request includes a design demand; extracting the design demand from the request; generating recommendation information on the basis of the design demand and a preset neural network of the ceramic design, wherein the recommendation information comprises a ceramic design scheme corresponding to the design demand. Through the recommendation method and device of the ceramic design, automatic recommendation is achieved, efficiency is improved, and moreover, since the recommendation information is generated in the neural network of the ceramic design on the basis of the design command, the efficiency is improved, and meanwhile, the recommendation effect is ensured, labor of a user is decreased, and the usage experience of the user is ensured.

The invention discloses a related patch recommendation method based on heterogeneous data. The method comprises the following steps: crawling multivariate heterogeneous data of code review; cleaning data, splicing multivariate heterogeneous data features into patch feature vectors; pairing patch pairs, taking the correlation with the prediction patch as a positive sample; taking a negative samplewithout correlation with the prediction patch; marking binary classification labels for the positive sample and the negative sample; dividing a training set and a verification set, using the trainingset to train a logistic regression model, a random forest model and a LightGBM model respectively to obtain a corresponding probability and a prediction label, calculating a corresponding accuracy rate according to the prediction label, and finally constructing a prediction score according to weighted summation of the fusion weight and the corresponding probability to obtain an optimal predictionscore. According to the method, correlation evaluation is carried out on the data submitted to the code review system through machine learning to obtain optimal recommendation, recommendation reliability and stability are improved, and more labor cost is saved.

The invention relates to a configuration parameter recommendation method and system, an instrument and a storage medium. The method comprises the following steps of acquiring instrument description information of a target instrument to be configured; constructing a knowledge graph of the target instrument according to the instrument description information; based on the knowledge graph of the target instrument and a stored historical instrument knowledge graph, determining a historical instrument matched with the target instrument; enabling the historical instrument and the target instrument to be the same type of instruments; and recommending configuration parameters corresponding to the historical instrument as configuration parameters of the target instrument. By adopting the method, the technical threshold and the debugging workload of technicians can be reduced.